refactor(data_source): default to mbp10 across smoke + localdev profiles

Smoke and localdev profiles previously used data_source = "ohlcv", divergent
from production (mbp10). Mismatch silently produced cache-key collisions in
the SHA256 hash path: smoke fxcache could not be loaded by production-shape
training without an explicit override. Local-dev fxcache regen also required
remembering to pass --data-source ohlcv to match.

Globalize mbp10 as the default everywhere it isn't deliberately overridden:
- config/training/dqn-smoketest.toml: data_source = "mbp10"
- config/training/dqn-localdev.toml: data_source = "mbp10"
- training_profile.rs: doc Default → "mbp10"
- trainers/dqn/config.rs: Default impl → "mbp10"
- hyperopt/adapters/dqn.rs: default → "mbp10"
- examples/precompute_features.rs: doc updated
- fxcache.rs / feature_cache.rs: discover_and_load + cache-key tests
  use "mbp10" arguments
- docs/dqn-wire-up-audit.md: new entry per Invariant 7

Documentation strings retained "ohlcv" only where they document the two
available choices (config.rs:946, training_profile.rs:83).

Local fxcache regenerated to v6 mbp10:
test_data/feature-cache/13c0b086a975cc7e2384377a2cd0e97738c9410292fcfecb5807c29bf885cb48.fxcache
(175874 bars, 55 MB, OFI_DIM=32). Stale v5 ohlcv fxcache untracked
from git index (already gitignored post-79578bbaf).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

config/training/dqn-localdev.toml

@@ -16,8 +16,8 @@ reward_scale = 1.0
 huber_delta = 1.0
 lr_decay_type = 2
 lr_min = 0.00001
-# Data source: "ohlcv" for local dev (may not have MBP-10 data)
-data_source = "ohlcv"
+# Data source: "mbp10" — matches production (MBP-10 imbalance bars)
+data_source = "mbp10"
 # OFI feature enrichment from MBP-10 order book data (8 features: OFI L1/L5, depth imbalance, VPIN, Kyle's lambda, bid/ask slope, trade imbalance)
 mbp10_data_dir = "test_data/futures-baseline-mbp10"
 trades_data_dir = "test_data/futures-baseline-trades"

config/training/dqn-smoketest.toml

@@ -22,7 +22,7 @@ reward_scale = 1.0
 huber_delta = 1.0
 symbol = "ES.FUT"
 max_bars = 5000
-data_source = "ohlcv"
+data_source = "mbp10"
 mbp10_data_dir = "test_data/futures-baseline-mbp10"
 trades_data_dir = "test_data/futures-baseline-trades"
 imbalance_bar_threshold = 1.0

crates/ml/examples/precompute_features.rs

@@ -133,20 +133,17 @@ struct Opts {
     ///
     /// MUST match the `data_source` field of the training profile that will
     /// consume this fxcache (e.g. `dqn-production.toml: data_source = "mbp10"`,
-    /// `dqn-smoketest.toml: data_source = "ohlcv"`). Mismatch produces a
+    /// `dqn-smoketest.toml: data_source = "mbp10"`). Mismatch produces a
     /// silent cache MISS at training time → DBN-direct fallback uploads
     /// un-normalised features → aux-head label_scale picks up raw-price
     /// magnitudes → cascade of broken metrics + impossible Sharpe (see
     /// 2026-04-27 incident: train-h5gxb epoch-0 Sharpe=141 from this exact
     /// path mismatch).
     ///
-    /// Defaults to `"mbp10"` because production (`dqn-production.toml`) is
-    /// the canonical consumer and uses MBP-10 microstructure data. Smoke
-    /// callers that intentionally use `dqn-smoketest.toml` (data_source =
-    /// "ohlcv") DO NOT call this binary — they reuse a pre-built local
-    /// fxcache committed under `test_data/feature-cache/`. If a smoke or
-    /// development workflow ever needs to (re)generate a local "ohlcv"
-    /// fxcache, pass `--data-source ohlcv` explicitly.
+    /// Defaults to `"mbp10"` — production, smoke, and localdev profiles all
+    /// use MBP-10 microstructure data. Pass `--data-source ohlcv` only when
+    /// intentionally generating a 1-minute candle cache for a profile that
+    /// overrides `data_source = "ohlcv"`.
     #[arg(long, default_value = "mbp10")]
     data_source: String,
 

crates/ml/src/feature_cache.rs

@@ -14,7 +14,7 @@ use std::path::{Path, PathBuf};
 /// by (canonical path, size, mtime). Any change to the directory contents
 /// (add / remove / modify) produces a different key.
 pub fn calculate_dbn_cache_key(data_dir: &Path) -> Result<String> {
-    calculate_dbn_cache_key_full(data_dir, None, None, "ES.FUT", "ohlcv")
+    calculate_dbn_cache_key_full(data_dir, None, None, "ES.FUT", "mbp10")
 }
 
 /// Extended cache key that includes MBP-10 and trades directories.
@@ -82,8 +82,8 @@ mod tests {
     fn test_cache_key_includes_symbol() {
         let dir = Path::new("test_data/futures-baseline");
         if !dir.exists() { return; }
-        let key_es = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "ohlcv").unwrap();
-        let key_nq = calculate_dbn_cache_key_full(dir, None, None, "NQ.FUT", "ohlcv").unwrap();
+        let key_es = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "mbp10").unwrap();
+        let key_nq = calculate_dbn_cache_key_full(dir, None, None, "NQ.FUT", "mbp10").unwrap();
         assert_ne!(key_es, key_nq, "Different symbols must produce different cache keys");
     }
 
@@ -91,7 +91,7 @@ mod tests {
     fn test_cache_key_includes_data_source() {
         let dir = Path::new("test_data/futures-baseline");
         if !dir.exists() { return; }
-        let key_ohlcv = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "ohlcv").unwrap();
+        let key_ohlcv = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "mbp10").unwrap();
         let key_mbp10 = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "mbp10").unwrap();
         assert_ne!(key_ohlcv, key_mbp10, "Different data sources must produce different cache keys");
     }
@@ -100,8 +100,8 @@ mod tests {
     fn test_cache_key_stable() {
         let dir = Path::new("test_data/futures-baseline");
         if !dir.exists() { return; }
-        let key1 = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "ohlcv").unwrap();
-        let key2 = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "ohlcv").unwrap();
+        let key1 = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "mbp10").unwrap();
+        let key2 = calculate_dbn_cache_key_full(dir, None, None, "ES.FUT", "mbp10").unwrap();
         assert_eq!(key1, key2, "Same inputs must produce same key");
     }
 }

crates/ml/src/fxcache.rs

@@ -610,7 +610,7 @@ pub fn norm_stats_path_for_key(
 /// * `symbol` — Trading symbol (e.g. `"ES.FUT"`)
 /// * `mbp10_dir` — Optional MBP-10 order book data directory
 /// * `trades_dir` — Optional trades data directory
-/// * `data_source` — Data source mode (`"ohlcv"` or `"mbp10"`)
+/// * `data_source` — Data source mode (`"mbp10"` or `"mbp10"`)
 /// * `cache_dir_override` — Explicit cache directory (from CLI `--feature-cache-dir`)
 pub fn discover_and_load(
     data_dir: &Path,
@@ -662,7 +662,7 @@ mod discover_tests {
             "ES.FUT",
             None,
             None,
-            "ohlcv",
+            "mbp10",
             None,
         );
         assert!(result.is_none());
@@ -691,7 +691,7 @@ mod discover_tests {
             "ES.FUT",
             None,
             None,
-            "ohlcv",
+            "mbp10",
             Some(&cache_dir),
         );
         // Strict match only — no "most recent" fallback

crates/ml/src/hyperopt/adapters/dqn.rs

@@ -545,7 +545,7 @@ pub struct DQNTrainer {
     /// Trading instrument symbol (e.g. "ES.FUT") — used in cache key to prevent
     /// cross-instrument collisions when multiple symbols share the same data directory.
     symbol: String,
-    /// Data source mode (e.g. "ohlcv", "mbp10") — used in cache key to prevent
+    /// Data source mode (e.g. "mbp10", "mbp10") — used in cache key to prevent
     /// collisions between different bar types built from the same files.
     data_source: String,
     /// TOML training profile name for per-trial hyperparameters (default: "dqn-production").
@@ -675,7 +675,7 @@ impl DQNTrainer {
             trades_data_dir: None,           // No trades (VPIN/Kyle) by default
             preloaded_ofi_features: None,    // Loaded during preload_data()
             symbol: "ES.FUT".to_string(),    // Default: ES futures
-            data_source: "ohlcv".to_string(), // Default: OHLCV bars
+            data_source: "mbp10".to_string(), // Default: MBP-10 imbalance bars
             training_profile: "dqn-production".to_string(), // Default: production params
         })
     }

crates/ml/src/trainers/dqn/config.rs

@@ -945,7 +945,7 @@ pub struct DQNHyperparameters {
 
     /// Data source for training bars: "ohlcv" (1-minute candles) or "mbp10"
     /// (imbalance bars from MBP-10 order book). No fallback -- fails if the
-    /// chosen source's data doesn't exist. Default: "ohlcv" (backward compatible).
+    /// chosen source's data doesn't exist. Default: "mbp10".
     pub data_source: String,
 
     /// Trading symbol subdirectory filter (e.g. "ES.FUT", "NQ.FUT").
@@ -1533,8 +1533,8 @@ impl DQNHyperparameters {
             mbp10_data_dir: "test_data/futures-baseline-mbp10".to_string(),
             trades_data_dir: "test_data/futures-baseline-trades".to_string(),
 
-            // Data source: "ohlcv" (backward compatible) or "mbp10" (imbalance bars)
-            data_source: "ohlcv".to_string(),
+            // Data source: "ohlcv" (1-min candles) or "mbp10" (imbalance bars from MBP-10 order book)
+            data_source: "mbp10".to_string(),
             symbol: "ES.FUT".to_string(),
             max_bars: 0, // 0 = unlimited
             imbalance_bar_threshold: 100.0,

crates/ml/src/training_profile.rs

@@ -81,7 +81,7 @@ pub struct TrainingSection {
     pub trades_data_dir: Option<String>,
 
     /// Data source for training bars: "ohlcv" (1-minute candles) or "mbp10"
-    /// (imbalance bars from MBP-10 order book). No fallback. Default: "ohlcv".
+    /// (imbalance bars from MBP-10 order book). No fallback. Default: "mbp10".
     pub data_source: Option<String>,
 
     /// Trading symbol subdirectory (e.g. "ES.FUT"). Filters DBN loading to one instrument.
@@ -786,7 +786,7 @@ impl DqnTrainingProfile {
             if let Some(ref v) = t.trades_data_dir {
                 hp.trades_data_dir = v.clone();
             }
-            // Data source selection: "ohlcv" or "mbp10"
+            // Data source selection: "mbp10" or "mbp10"
             if let Some(ref v) = t.data_source {
                 hp.data_source = v.clone();
             }

docs/dqn-wire-up-audit.md

@@ -2,6 +2,27 @@
 
 **Status:** Populated during Plan 1 Task 6 (A.5 orphan audit). Updated on every commit per Invariant 7.
 
+data_source globalization to mbp10 (2026-04-27): completes the alignment
+started in the earlier "fxcache data_source alignment" entry below. Smoke
+(`dqn-smoketest.toml`) and localdev (`dqn-localdev.toml`) profiles
+previously set `data_source = "ohlcv"` while production sets `"mbp10"`.
+The split forced two separate fxcache files and made cross-environment
+runs require explicit `--data-source ohlcv` overrides on the precompute
+binary. Default rolls forward to `"mbp10"` everywhere it is not
+deliberately overridden:
+- `config/training/dqn-smoketest.toml`, `dqn-localdev.toml`: profile values
+- `crates/ml/src/training_profile.rs`, `trainers/dqn/config.rs`,
+  `hyperopt/adapters/dqn.rs`: Rust defaults + doc strings
+- `crates/ml/examples/precompute_features.rs`: doc updated;
+  `--data-source mbp10` is the canonical regen invocation
+- `crates/ml/src/feature_cache.rs`, `fxcache.rs` test fixtures use mbp10
+Documentation references to the alternative ("ohlcv") are retained only
+where the docstring explicitly enumerates the two valid choices. Local
+fxcache regenerated to v6 mbp10
+(`13c0b086a975cc7e2384377a2cd0e97738c9410292fcfecb5807c29bf885cb48.fxcache`,
+55 MB, 175874 bars). Stale v5 ohlcv cache untracked from git index
+(already gitignored post-`79578bbaf`).
+
 MoE Phase 3 wire-up T3.1–T3.7 (2026-04-27): MoE fully wired into
 production training path. Forward: gate (state[B,128]→64→8→softmax) +
 8 expert MLPs (h_s1[B,256]→64→256) + `moe_mixture_forward` → replaces